Enzyme AD

Enzyme Automatic Differentiation Framework

Enzyme Conference 2023


The first Enzyme workshop took place from February 22-24 in Boulder, Co. The workshop’s aim was to bring together the contributors and users of Enzyme in particular, and people interested in automatic differentiation within the LLVM infrastructure in general, for two days of fast-paced high-bandwidth idea exchange and collaboration bootstrapping. For this the workshop was designed to have a flexible format, friendly to practitioners and researchers alike; its program consisted of a mix of presentations, tutorials, roundtable discussions and live programming sessions.


Session 1: Enzyme Applications 

  • PDSR-Enzyme: Physics-Based Differentiable Rendering Using Enzyme; Shuang Zhao, Zihan Yu, Cheng Zhang
  • Automatic Differentiation in Solid Mechanics: Interpretation and Composition; Jed Brown, Leila Ghaffari
  • Differentiating Large-Scale Finite Element Applications with MFEM; Julian Andrej
  • A Cross-Language Probabilistic Programming Protocol for Physics and Beyond; Tim Gymnich, Ludger Paehler
  • Numba-Enzyme: A Differentiable JIT-ed Python; Ludger Paehler, Jan Hueckelheim, Nikolaus A. Adams, Lukas Heinrich

Session 2: Lessons from AD Tools 

  • A survey of Tapenade in contrast with Enzyme; Laurent Hascoet
  • NIFTy: The Why and How of Building AD from Scratch; Gordian Edenhofer
  • AD with Integrals; Jesse Michel

Session 3: Enzyme in Modern Programming Languages 

  • Compilation Augmentation Enables High-Performance Batch Differentiation; Tim Gymnich
  • Differentiable molecular simulation with Molly.jl; Joe Greener
  • Adjoint Checkpointing using Custom Differentiation Rules; Michel Schanen
  • Oxide-Enzyme - Integrating Enzyme into the Rust Compiler; Manuel Drehwald, Lorenz Schmidt

Session 4: Enzyme Extensions 

  • Hackable Autodiff: Extending Enzyme to MLIR for Reverse Mode Gradients; Martin Eppert, Jacob Mai Peng
  • Autodiff semantics and the level of abstraction; Jan Hueckelheim
  • Accelerating Black Hole Imaging with Enzyme; Paul Tiede


  • DJ4Earth: Oceans, Ice Sheets, Adjoints, and AD; Sarah Williamson, Patrick Heimbach